Generate automated test instructions for your software or application features using images and optional context with the power of Hugging Face's ViLT model!
This project provides a web-based interface to generate test instructions using Visual Question Answering (VQA). Upload screenshots or images, provide an optional context, and let the model do the rest!
- Visual Question Answering: Automatically generate test instructions based on uploaded images.
- Context-Aware: Include optional text context to refine the instructions.
- Dynamic Interface: Simple and user-friendly web interface for easy interaction.
- Scalable: Easily extendable with more models or features.
├── app/
│ ├── app.py # Main Flask application
│ ├── uploads/ # Directory for uploaded images
│ └── utils/
│ └── llm.py # LLM utilities (model inference functions)
├── templates/
│ └── index.html # HTML template for the web form
├── static/
│ └── css/
│ └── styles.css # Custom CSS for styling
├── README.md # Project documentation
├── requirements.txt # Python dependencies
└── .gitignore # Files and directories to ignore in Git
1.) Python 3.7+: Ensure you have Python installed.
2.) Pip: Make sure pip is installed to manage dependencies.
- Clone the Repository:
git clone https://github.com/Blacksujit/TestCraftAI.git
cd app
- Create a Virtual Environment:
python -m venv venv
On Windows: venv\Scripts\activate
- Install Dependencies:
pip install -r requirements.txt
- Download the Model:
transformers-cli download dandelin/vilt-b32-finetuned-vqa
- Run the Application:
python app.py
- Access the Application:
Open your web browser and go to: http://127.0.0.1:5000
1.) Upload Images: Click on the upload button and select the images/screenshots you want to use for generating test instructions.
2.) Add Context: (Optional) Provide context to make the instructions more specific.
3.) Generate Instructions: Click the "Generate" button to receive your test instructions.
Feel free to contribute by forking the repository, making changes, and submitting a pull request!
1.) Add support for multiple languages.
2.) Integrate more advanced models.
3.) Improve UI/UX for better interaction.
Contributions are welcome! Please read the CONTRIBUTING.md for more details.
This project is licensed under the MIT License - see the LICENSE file for details.
1.) Hugging Face for the ViLT model.
2.) Flask for providing the web framework.
3.) Community contributors for their continuous support.